k4yt3x/video2x
A machine learning-based video super resolution and frame interpolation framework. Est. Hack the Valley II, 2018.
Built in C/C++ for performance, the framework integrates multiple upscaling engines (Real-ESRGAN, Real-CUGAN, Anime4K v4) and frame interpolation (RIFE) via ncnn and Vulkan backends, supporting custom MPV-compatible GLSL shaders. Processes video with zero intermediate disk overhead through optimized GPU-accelerated pipelines, requiring only AVX2 CPUs and Vulkan-capable GPUs. Deployable via Qt6 GUI, CLI, Docker containers, or Google Colab for accessibility across Windows, Linux, and cloud environments.
19,070 stars. Actively maintained with 2 commits in the last 30 days.
Stars
19,070
Forks
1,676
Language
C++
License
AGPL-3.0
Category
Last pushed
Mar 07, 2026
Commits (30d)
2
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/computer-vision/k4yt3x/video2x"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related tools
jhogsett/EMA-VFI-WebUI
Advanced AI-Based Video Renovation UI Using EMA-VFI & Real-ESRGAN
MediaTek-NeuroPilot/mai22-real-time-video-sr
The official codebase for the Real-Time Video Super-Resolution Challenge in Mobile AI (MAI)...
JihyongOh/XVFI
[ICCV 2021, Oral 3%] Official repository of XVFI
EutropicAI/ccrestoration
an inference lib for image/video restoration with VapourSynth support